from animosity to affinity: the interplay of competing ... · defined as “socially constructed,...
TRANSCRIPT
The University of Manchester Research
From Animosity to Affinity: The Interplay of CompetingLogics and Interdependence in CrossSector PartnershipsDOI:10.1111/joms.12273
Document VersionAccepted author manuscript
Link to publication record in Manchester Research Explorer
Citation for published version (APA):Ashraf, N., Pinkse, J., & Ahmadsimab, A. (2017). From Animosity to Affinity: The Interplay of Competing Logicsand Interdependence in CrossSector Partnerships. Journal of Management Studies, 54(6), 793-822.https://doi.org/10.1111/joms.12273
Published in:Journal of Management Studies
Citing this paperPlease note that where the full-text provided on Manchester Research Explorer is the Author Accepted Manuscriptor Proof version this may differ from the final Published version. If citing, it is advised that you check and use thepublisher's definitive version.
General rightsCopyright and moral rights for the publications made accessible in the Research Explorer are retained by theauthors and/or other copyright owners and it is a condition of accessing publications that users recognise andabide by the legal requirements associated with these rights.
Takedown policyIf you believe that this document breaches copyright please refer to the University of Manchester’s TakedownProcedures [http://man.ac.uk/04Y6Bo] or contact [email protected] providingrelevant details, so we can investigate your claim.
Download date:22. Mar. 2020
From Animosity to Affinity: The Interplay of Competing Logics and Interdependence in
Cross-Sector Partnerships
Naeem Ashraf Assistant Professor
Suleman Dawood School of Business
Lahore University of Management Sciences
Opposite Sector U, DHA, Lahore, Pakistan
&
Assistant Professor
Montpellier Business School
2300 Avenue des Moulins
34185 Montpellier Cedex 4, France
email: [email protected]
Alireza Ahmadsimab Assistant Professor
Sobey School of Business, Saint Mary’s University
923 Robie St, Halifax, NS, B3H 3C3, Canada
email: [email protected]
Jonatan Pinkse*
Full Professor
Alliance Manchester Business School, University of Manchester
Booth Street West
Manchester M15 6PB, United Kingdom
e-mail: [email protected]
* Corresponding author
This article has been accepted for publication and undergone full peer review but has not beenthrough the copyediting, typesetting, pagination and proofreading process which may lead todifferences between this version and the Version of Record. Please cite this article as a Journalof Management Studies (‘Accepted Article’), doi: 10.1111/joms.12273
This article is protected by copyright. All rights reserved.
2
Biographies
Naeem Ashraf is Assistant Professor of Strategy and Organization at Suleman Dawood School
of Business, Lahore University of Management Sciences, Pakistan, and at Montepellier Business
School, France. He obtained his PhD from IAE Aix-en-Provence, Aix-Marseille Université,
France. During his professional career that spans more than fifteen years, he served in senior
positions in different sectors that include fertilizers, automobile, space research, and higher
education. His research interests include social and inter-organizational networks, and corporate
sustainability. Previously he has published in Organization & Environment, and Scientometrics.
Alireza Ahmadsimab is Assistant Professor in the Management Department of Sobey School of
Business at Saint Mary’s University, Canada. He obtained his Ph.D. in business administration
from ESSEC Business School. His research lies at the intersection of strategy, organizational
theory and international business. In his research, he employs qualitative and quantitative
methods to study how organizations in Cross Sector Social Partnerships (CSSPs) manage to
establish stable strategic partnerships. In parallel with his research, he offers courses in strategic
management, international business and organization theory at the undergraduate, Masters and
MBA level.
Jonatan Pinkse is a Professor of Strategy, Innovation, and Entrepreneurship at the Manchester
Institute of Innovation Research, Alliance Manchester Business School, The University of
Manchester, UK. He obtained his PhD from the University of Amsterdam. Before moving to
Manchester he held positions at the University of Amsterdam and Grenoble Ecole de
Management. His research interests focus on corporate sustainability, business responses to
climate change, cross-sector partnerships, and sustainable business models and innovation. His
articles have appeared in such journals as Academy of Management Review, Entrepreneurship
Theory and Practice, Journal of International Business Studies, and Organization Studies.
This article is protected by copyright. All rights reserved.
3
ABSTRACT. Drawing on and extending institutional logics and resource dependence theories,
this paper posits that for cross-sector partnerships to survive, organizations need to share
compatible institutional logics, but depend less on each other’s resources. Asymmetrical cross-
sector partnerships may lead to a breakup if organizations are forced to operate under
incompatible institutional logics. The findings of this study show that the challenges posed by
incompatible logics of partners could be mitigated by the degree of resource interdependence
between organizations. Capturing the effects of context and transactions on the actors’ strategic
behavior, the findings, based on dataset of project-level partnership ties between 1312
organizations in the carbon-offset market, support these hypotheses. The paper concludes by
discussing implications of organizations' responses to keep acting under or reinterpreting existing
institutional logics in asymmetrical cross-sector relationships.
Keywords: Cross-sector partnerships, Resource dependence, Nonprofit organizations
Institutional complexity, Carbon market
This article is protected by copyright. All rights reserved.
4
INTRODUCTION
Organizations are increasingly partnering across sectors to solve various societal and
environmental problems such as health, education, environmental degradation, climate change,
and poverty alleviation (Teegen, Doh and Vachani, 2004; Selsky and Parker, 2005; Yaziji and
Doh, 2009). Cross-sector partnerships are initiatives where organizations from different sectors,
i.e. for-profit and non-profit sectors, enter into a collaboration to achieve a common objective
(Selsky and Parker, 2005). When resources in one sector are insufficient to solve these problems,
a cross-sector partnership may be a viable option (Austin and Seitanidi, 2012; Kivleniece and
Quelin, 2012). However, cross-sector partnerships have often led to conflict-laden relationships
and are frequently susceptible to failure (Selsky and Parker, 2005; Prashant and Harbir, 2009).
Research examining difficulties in the maintenance of cross-sector partnerships has
identified several factors that influence the survival of partnerships, including the frequency of
contact between partners, the organizations’ level of commitment to socially responsible
behavior, and the existence of a personalized, reciprocal bond between organizations
(Drumwright, Cunningham and Berger, 2004; Le Ber and Branzei, 2010). Such partnerships also
face problems similar to those of alliances between firms, such as opportunistic behavior of
partners, poor communication and coordination, and differences in management structures (Park
and Ungson, 2001). Nonetheless, we still know relatively little about the determinants of failure
that are more specific to cross-sector partnerships. In this paper, we argue that a distinctive
source of failure of cross-sector partnerships is associated with how compatible the partners are
regarding their cultural backgrounds and the objectives they seek. Organizations in cross-sector
partnerships usually pursue different and potentially conflicting objectives (Austin, 2000). The
different objectives that cross-sector partners pursue are not only a result of organizational-level
This article is protected by copyright. All rights reserved.
5
decisions, but also of broader cultural templates, or institutional logics, which infuse
organizations with specific values (Thornton, Ocasio and Lounsbury, 2012). That is,
“organizational members, by being part of social and occupational groups, enact, within
organizations, broader institutional logics that define what actors understand to be the
appropriate goals, as well as the appropriate means to achieve these goals” (Pache and Santos,
2010: 459).
We propose that employing the institutional logics perspective will shed light on factors
that might undermine the survival of cross-sector partnerships. Institutional logics have been
defined as “socially constructed, historical patterns of material practices, assumptions, values,
beliefs, and rules by which individuals produce and reproduce their material subsistence,
organize time and space, and provide meaning to their social reality” (Thornton and Ocasio,
1999: 804). Cross-sector partnerships bring together organizations that are embedded in multiple,
often competing and overlapping institutional logics – a situation referred to as institutional
complexity (Greenwood et al., 2011). An excessive rivalry between institutional logics can cause
conflict and instability; however, few studies have examined the impact of incompatible
institutional logics on the survival of cross-sector partnerships (Vurro, Dacin and Perrini, 2010).
While cross-sector partnerships might suffer from an incompatibility of logics, we argue
that a sole focus on this incompatibility does not provide the full picture in explaining
partnership survival. Partnerships tend to be initiated from a need to exchange critical resources
to tackle societal and environmental problems (Austin and Seitanidi, 2012). The need to acquire
resources has a solid footing in the inter-organizational relationship literature (Pfeffer and
Salancik, 1978; Casciaro and Piskorski, 2005). Studies have explored the role of
interdependence and its influence on inter-organizational partnership dynamics (Oliver, 1991;
This article is protected by copyright. All rights reserved.
6
Beckman, Haunschild and Phillips, 2004; Prashant and Harbir, 2009). Since different
organizations in a partnership share resources and depend on each other, they have the interest
to continue the partnership even if they face obstacles (Gulati and Sytch, 2007; Wry, Cobb and
Aldrich, 2013). This interdependence raises the issue of how organizations in a cross-sector
partnership deal with the tension between the need to exchange resources, on the one hand, and
cooperate with organizations whose behaviors are guided by different institutional logics, on the
other (Wry et al., 2013). This study addresses this underexplored issue by investigating the
combined impact of institutional complexity and resource dependency on the survival of cross-
sector partnerships.
As an empirical setting, we selected the carbon-offset market – the Clean Development
Mechanism (CDM) – that operates under the Kyoto Protocol. CDM allows organizations from
industrialized countries to reduce greenhouse gas (GHG) emissions by taking part in projects in
developing countries. A CDM project tends to be organized as a cross-sector partnership (Pinkse
and Kolk, 2012), bringing together for-profit organizations, NGOs and/or governments from
different countries in different constellations. The CDM market provides a suitable empirical
setting for testing our hypotheses on the influence of incompatible logics and resource
interdependency on partnership survival because the partnering organizations have rather
distinctive views on the purpose of projects. Nonetheless, partners have a clear incentive to
collaborate and bridge differences as they share the need of showing that projects are legitimate
to get them financed (Pinkse and Kolk, 2012). To test our hypotheses, we compiled an original
dataset consisting of partnerships between 1312 organizations from 59 countries working on
1500 carbon-offset projects during the observation years 2005 to 2010.
This article is protected by copyright. All rights reserved.
7
Theoretically, this paper offers a new understanding of factors that might destabilize
cross-sector partnerships by highlighting two potential sources of tension: an incompatibility of
institutional logics and the level and kind of organizational interdependence. The study’s
findings suggest that incompatible institutional logics negatively affect the survival of cross-
sector partnerships, but interdependence between them moderates this relationship. Building on
Casciaro and Piskorski (2005), we study the effects of two components of interdependence:
mutual dependence and power imbalance. While mutual dependence positively moderates the
relationship between incompatible logics and partnership survival, we found the opposite for
power imbalance between partners as it intensifies the adverse impact of incompatible logics on
partnership survival. This study further contributes to the growing body of research on
institutional complexity (Greenwood et al., 2011) by bringing resource dependence to the
forefront of the analysis (Wry et al., 2013). Research investigating organizational responses to
conflicting institutional demands tends to turn a blind eye to the role of resource
interdependence (Reay and Hinings, 2005; Townley, 2002; Dunn and Jones 2010; Battilana and
Dorado, 2010; Wry et al., 2013). Empirically, this study is among the few that investigate the
impact of competing institutional logics by focusing on the survival of cross-sector partnerships.
By integrating elements from institutional logics and resource dependence theory, the study
provides new insights into understanding what factors undermine cross-sector partnerships.
THEORY AND HYPOTHESES
Incompatible Institutional Logics in Cross-Sector Partnerships
Cross-sector partnerships constitute a subject of growing importance in management
research (Austin and Seitanidi, 2012; Prashant and Harbir, 2009; Reed and Reed, 2009). These
This article is protected by copyright. All rights reserved.
8
partnerships tend to utilize the for-profit sector’s capacity in wealth creation to address social
issues such as environmental sustainability, economic development, health, education, and
poverty alleviation (Teegen et al., 2004; Selsky and Parker, 2005; Yaziji and Doh, 2009). As
resources in the non-profit sector (comprising governments and non-governmental organizations)
are insufficient to respond to such issues, the for-profit sector is increasingly called upon to share
financial and informational resources (Rangan et al., 2006; Pinkse and Kolk, 2012). For-profit
organizations (FPOs) are motivated to contribute to social issues and engage with non-profit
organizations (NPOs), responding to societal pressure to go beyond profit maximization and
shareholder value and pursue broader social benefits for their stakeholders (Marquis et al., 2007).
Previous studies have identified mechanisms through which value is created and
sustained in cross-sector partnerships (Le Ber and Branzei, 2010; Kivleniece and Quelin, 2012).
Nonetheless, there are indications that cross-sector partnerships are fragile (Teegen et al., 2004;
Drumwright et al., 2004). A distinctive tension in cross-sector partnerships is that the practices
of for-profit and non-profit organizations tend to bear the imprint of different institutional logics
(Vurro et al., 2010). Institutional logics guide organizational behavior by providing specific
scripts for action, establishing core principles for organizing activities, and channeling interests
(Thornton et al., 2012). However, cross-sector partnerships bring together logics that are not
necessarily compatible. Compatibility of logics refers to the extent to which logics provide
congruous prescriptions for action (Besharov and Smith, 2014; Greenwood et al., 2011). As
Besharov and Smith argue (2014: 369), “logics are more compatible when they provide
consistent and reinforcing prescriptions for actions and belief.” Incompatible logics can thus be
a source of conflict between organizations in cross-sector partnerships (Vurro et al., 2010) and
threaten performance, ultimately leading to partnership dissolution.
This article is protected by copyright. All rights reserved.
9
Previous studies indicate that an important incompatibility in cross-sector partnerships
relates to the clash between a public good logic and a market logic (Mars and Lounsbury, 2009;
Battilana and Dorado, 2010). Cross-sector partnerships tend to be set up to provide a public
good of addressing a social or environmental problem. However, partners also have their private
interests as they wish to appropriate part of the value that is created in the partnership (Di
Domenico, Tracey and Haugh, 2009; Hahn and Pinkse, 2014; Kivleniece and Quelin, 2012;
Rangan et al., 2006). The market logic and ensuing need to appropriate private benefits from
cross-sector partnership collaboration are most obviously present in FPOs (i.e., private firms
and publicly-listed state-owned firms) because maximizing shareholder value is their primary
purpose. It has been argued that this purpose also has a legal dimension. FPOs have a fiduciary
duty to shareholders to prioritize their interests over those of other stakeholders (Friedman,
1970; Jensen, 2001). While this narrow shareholder view has been contested, it is still pervasive
in the for-profit sector (Reinhardt, Stavins and Vietor, 2008). In a partnership context,
Kivleniece and Quelin (2012: 277) argued, for example, that “[f]irms are expected to seek and
capture not only financial returns (rent) from exploiting an underlying business opportunity but
also a range of intangible assets and institutional benefits, such as access to public knowledge
sources, enhanced organizational skills and routines, and reputation or legitimacy spillovers
from an engagement in domains of public interest.” In contrast, the public good logic tends to
be more closely associated with NPOs (i.e., governments and non-governmental organizations),
as their main aim is to “strive to maximize public benefits (indicated, say, as region-level GDP
per capita and sustainable development) for their constituents” (Rangan et al., 2006: 741). NPOs
will have fewer issues with not being able to appropriate a partnership’s benefits accruing to
society. The legal mandate of government agencies is to pursue multiple social welfare
This article is protected by copyright. All rights reserved.
10
objectives besides the economic dimension (La Porta et al., 1999), while NGOs’ legal
accountability is inherent in the nature of their organizations and the causes and people they
serve (Candler and Dumont, 2010), including clients, experts, staff, donors, regulators,
members, the general public and the media (Brown and Moore, 2001).
We expect cross-sector partnerships to suffer from a lack of compatibility between the
market logic and the public good logic, which could threaten partnership survival. We assume
that organizations from each sector will be driven by a dominant logic (Reay and Hinings,
2009). From an institutional point of view, organizations are not monolithic and have to balance
competing institutional demands (Pache and Santos, 2010). Moreover, the distinction between
sectors is becoming more blurred: government functions are being delegated to the private
sector (Kivleniece and Quelin, 2012); firms move towards the public sector by being hybrid
organizations (Pache and Santos, 2010; Battilana and Dorado, 2010); and NGOs increasingly
follow the social enterprise model, using for-profit tactics to increase impact (Teegen et al.,
2004). Nonetheless, we expect the clash between the public good and market logics to be
substantial in cross-sector partnerships. Most FPOs and NPOs are still worlds apart in terms of
the logics that guide their behavior; they usually have little knowledge about each other’s
operations, culture, and values and have different expectations from engaging in a partnership
(Drumwright et al., 2004). Previous research indicates that such differences between FPOs and
NPOs affect their exchange relationship and can make collaboration challenging (Teegen et al.,
2004; Drumwright et al., 2004). An incompatibility of logics between partners would thus have
an adverse impact on partnership survival.
This article is protected by copyright. All rights reserved.
11
Hypothesis 1: The higher the level of incompatibility between the institutional
logics present in a cross-sector partnership, the higher the hazard of dissolution of
the partnership.
Resource Dependencies in Cross-Sector Partnerships
While incompatible logics pose challenges for cross-sector partnerships, we expect this
effect to be mitigated by the degree of resource interdependence between organizations.
Notwithstanding the influence of logics, resource dependence theory argues instead that
organizational practices tend to follow more rational patterns of decision-making (Oliver, 1991;
Binder, 2007). According to resource dependence theory, organizations engage in partnerships
to access resources to achieve objectives that they are unable to achieve on their own (Pfeffer
and Salancik, 1978). Interdependence is defined as the degree of partners’ need for each other’s
resources to achieve their respective goals. Organizations establish partnerships to obtain
resources (physical and/or symbolic) that are not available from their own environments (Katila,
Rosenberger and Eisenhardt, 2008). The literature on cross-sector partnerships has noted
various opportunities for sharing organizational resources. FPOs have deep financial pockets
and commercial skills, but NPOs have expertise regarding the generation and distribution of
public goods and have negotiation and facilitation skills. While NPOs look for financial or
human resources in partnerships (Brown and Kalegaonkar, 2002; Sagawa and Segal, 2000),
FPOs seek social reputation and legitimacy instead (Austin, 2000; Dacin et al., 2007).
Intuitively, we would expect interdependence to lessen the adverse impact of incompatible
logics on partnership survival. The more cross-sector partners depend on each other for
This article is protected by copyright. All rights reserved.
12
resources, the more they share an interest to keep the partnership alive, even if this involves
dealing with conflicts from incompatible logics (Wry et al., 2013).
Drawing on recent work in resource dependence theory (Casciaro and Piskorski, 2005;
Gulati and Sytch, 2007), we distinguish between a balance and size effect of interdependence to
provide a more nuanced picture of its moderating role. The balance effect refers to power
differentials between two organizations in a partnership. When the dependence relation is not
purely symmetric, one of the partners wields more power (Casciaro and Piskorski, 2005). The
size effect refers to the magnitude of the dependencies. High interdependence occurs when
resources of both partners are essential for achieving the goals of the partnership. Low
interdependence refers to a situation where partners’ resources are desired, but not crucial for
the partnership’s success, or easily replaceable. To disentangle the balance and size effect, in
their reformulation of resource dependence theory, Casciaro and Piskorski (2005) introduced
two dimensions of resource dependence. The first dimension, ‘power imbalance’, is the extent
to which partners depend critically on the other or have alternative options available to access
similar resources. The second dimension, ‘mutual dependence’, captures bilateral dependencies
regardless of whether the partners’ dependencies are balanced or imbalanced.
Regarding the moderating effect of power imbalance, as the interdependence might entail
an asymmetric power relation (Casciaro and Piskorski, 2005), the more powerful partner would
attempt to exercise its power over the other by imposing its institutional logic. This situation has
been referred to as a ‘substitution of logics’ where one of the logics gains dominance over the
other (Reay and Hinings, 2005; Townley, 2002). A substitution scenario means the triumph of
the institutional logic of the powerful organization and the eradication of the logic of the other.
The latter will accept the dominance of the competing logic only because it needs the former’s
This article is protected by copyright. All rights reserved.
13
resources to achieve an objective. We expect that whether organizations see a power imbalance
as a reason to dissolve the partnership depends on the degree to which they fear ‘mission drift’
to occur (Moore, 2000; Margolis and Walsh, 2003). If engaging in a cross-sector partnership
could entail a full substitution of logics, the risk for an organization of failing to stay true to its
mission would be too high. For example, FPOs might be reluctant to participate in cross-sector
partnerships that do not create economic value, because involvement in such ‘purely’ social
activities may yield results other than shareholder value (Margolis and Walsh, 2003). NPOs
might also fear that by establishing a partnership with business, they move away from their
fundamental values such as ensuring the respect for human rights and lose legitimacy in the eye
of their supporters (Rondinelli and London, 2003). Institutional logics run deep into the
organizations’ culture, such that any incongruent change in logics is resisted, and partnership
failure may be opted for, instead of compromising on the logics. Despite the need for resources,
organizations may prefer to give up a partnership that threatens their dominant logic. The more
the logics of partners are incompatible, the more likely they will see a power imbalance as a
reason to abandon the partnership. Therefore, we expect power imbalance to aggravate the
impact of incompatible logics on the survival of cross-sector partnerships.
Hypothesis 2: The higher the power imbalance between cross-sector partners,
the higher the impact will be of incompatible logics on the hazard of
dissolution of a partnership.
Regarding the moderating effect of mutual dependence, we expect the opposite to occur.
When the value created by cross-sector partners by working together is high – a situation of
This article is protected by copyright. All rights reserved.
14
mutual dependence (Casciaro and Piskorski, 2005) – partners have reasons to keep working
together, a scenario referred to as a ‘co-existence of logics’ (Kraatz and Block, 2008; Dunn and
Jones 2010; Reay and Hinings, 2009). In the case of highly incompatible institutional logics,
partners challenge and confront each other’s assumptions, norms and values, and practices, but
the need for critical resources leads them to adapt, even if they oppose each other’s logic
(DiMaggio and Powell, 1983). Although highly incompatible logics will cause tension in the
partnership, high mutual dependence provides enough incentives to organizations to find
solutions that reconcile the different logics. If all partners attempt to emphasize their autonomy
on norms, values, and organizational practices, the success of the partnership will not be
sustainable (Kaplan, 2008); success requires ongoing compromise (Croteau and Hicks, 2003).
We expect that whether organizations see mutual dependence as a reason to sustain the
partnership depends on the degree to which they consider the partnership as strategically
important (Austin, 2000; Rondinelli and London, 2003). In strategically important partnerships,
partners make higher resource commitments, leading to higher mutual dependence and a
considerable effort to reconcile competing logics. Previous studies have discussed mechanisms
that cross-sector partners use to mutually adjust their organizational routines and expectations
(Wang, Choi, and Li 2008; Katila et al., 2008; Seitanidi and Crane, 2009). For instance,
organizational members innovatively devise new organizational practices and actively reinterpret
the situation to frame it in their legitimate terms. Therefore, we expect mutual dependence to
alleviate the adverse impact of incompatible logics on the survival of cross-sector partnerships.
This article is protected by copyright. All rights reserved.
15
Hypothesis 3: The higher the mutual dependence between cross-sector
partners, the lower the impact will be of incompatible logics on the hazard of
dissolution of a partnership.
METHOD
Sample and Data
We selected the carbon-offset market that operates under the Kyoto Protocol’s Clean
Development Mechanism (CDM) as our empirical setting. CDM is a project-based market that
provides organizations with a cost-efficient way of meeting sustainable development targets
(Kolk and Mulder, 2011). Under CDM, organizations enter into partnerships to reduce GHG
emissions generated by investments in projects commissioned in a developing country. Some of
these projects last for 30 years. Currently, China and India are the major target countries for
CDM projects. The uncertainty and institutionalization of the emissions trading regime and the
existence of cross-sector partnerships between developed and developing countries’
organizations in this market make it an interesting empirical context to explore the interplay
between different institutional logics and resource dependencies. Moreover, there are
fundamentally different views on the purpose of CDM. While some see it as a cost-efficient way
to reduce emissions, others view it as a mechanism to foster sustainable development (Olsen and
Fenhann, 2008).
The carbon-offset project-level data was obtained from the Institute for Global
Environmental Strategies1 and triangulated with the official CDM database
2 as well as project
documents downloaded from this database. Information about the sample organizations was
obtained from Mint Global – a proprietary database3 of firm information. For Chinese
This article is protected by copyright. All rights reserved.
16
organizations, we used official statistical data published annually by the National Bureau of
Statistics of China4. We also obtained information from project documents and publicly available
information on the Internet.
We focused on those CDM projects, which earned carbon emission reduction credits
(CERs) during the 2005-2010 observation period. To establish when project participants started
or joined a project, we first referred to the original Project Design Document (PDD), which
mentions the organizations working on that project. The moment the PDD was submitted to the
United National Framework Convention on Climate Change (UNFCCC) secretariat was taken as
the date that partners officially became project participants. If a partner joined later and PDDs
did not mention it, we checked when the Designated National Authority (DNA), whose approval
is necessary, issued its authorization. We took this as the organization’s joining date. Following
Pattison et al. (2013), we then tracked the relationships of these actors with each other (first-tier
network) and their immediate partners (second-tier network), whose relationships with others
(third-tier network) were also included in our sample. The process was continued until we only
found redundant ties or loops with other organizations in the dataset. This data collection
resulted in a quasi-cohesive network of partnerships where organizations had more ties within
the network than outside.
The resulting sample consisted of 1312 organizations from 59 countries (representing 39%
of the population of organizations whose projects were registered in the CDM database). The
sample organizations were working on 1500 carbon-offset projects. Varying for each year, the
majority was either Chinese or Indian and in terms of the industry either from electrical services
or the cement industry. Indian organizations’ concentration in the sample decreased from 21% in
2005 to 10% in 2010 (see Table I for detailed statistics), while the concentration of Chinese
This article is protected by copyright. All rights reserved.
17
organizations increased. After reaching 30% in 2008, the share of Chinese organizations dropped
to about 19% in 2010, but still maintained a relatively larger share than Indian organizations. The
partnership ties between the sample organizations varied each year during the observation years,
reaching a maximum of 9041 project-level ties or 2899 cross-sector project-level ties. Cross-
sector partnership ties at the organizational level peaked at 1174 ties in 2009, after which the
number of ties dropped, thus indicating a process of tie dissolution. Most organizations in the
sample were for-profits, i.e. private and publicly listed state-owned firms, the latter of which held
a comparatively low proportion of renewable energy projects (see Table I for sample statistics).
-------------------------------------------
INSERT TABLE I ABOUT HERE
-------------------------------------------
Dependent Variable
The dependent variable in our study is the dissolution of partnership ties at the project
level. Through official modality of communication documents, organizations convey information
about any changes such as a withdrawal from a project. To establish when an organization
withdrew from a project and dissolved a tie, we took the date when this information was
submitted to the UNFCCC. Following previous approaches (Park and Russo, 1996; Polidoro,
Ahuja and Mitchell, 2011), a dummy variable - dissolution - was coded for each tie-dissolution
at the project level. The dummy was coded 1 if a tie was dissolved in that year and 0 otherwise.
Independent Variables
Incompatibility between institutional logics. To measure incompatibility of institutional
logics, we first considered a CDM project’s dominant logic by establishing the type of
technology it was based on, i.e. renewable or non-renewable. Notwithstanding the efforts made
This article is protected by copyright. All rights reserved.
18
to reduce GHG emissions, we argue that organizations whose CDM projects focus on non-
renewable technologies are still locked into the fossil-fuel paradigm. While they try to reduce
emissions, they fail to do this in a way that would transform the economy by moving away from
fossil fuels (Geels, 2010). Their incentive to participate in CDM projects is more closely aligned
with the market logic of maximizing private benefits from obtaining CERs at the lowest cost
possible, either for financial gain or to comply with regulatory pressure in their home country.
CDM projects that focus on renewable technologies not only reduce GHG emissions, but also
aim to transform the economy towards a higher uptake of sustainable sources of energy (Olsen
and Fenhann, 2008). This transformation is not only technological in nature but also institutional,
as it requires changes in “practices, policy, and cultural meanings” (Geels, 2010: 494). Hence,
organizations that invest relatively more in renewable energy projects are more closely aligned
with a public good logic as part of the benefit does not accrue to them but to the broader society.
In line with existing literature, CDM projects that were based on biomass, biogas,
hydropower, wind power, and other renewable energy technologies were classified as renewable
energy projects (Olsen and Fenhann 2008). All other projects were classified as non-renewable.
We coded renewable projects as 0 and non-renewable projects as 1, and for each organization
counted the number of projects in each category. Then, we computed the year-wise proportion of
renewable projects by dividing the number of renewable projects by the total number of projects
of each organization. To compute our main predictor – incompatibility – the difference in
proportion of renewable energy projects of each organization in a dyad was computed and
denoted as Ir:
�� = |�(�) − �()|
where � = � ����������������������������������� ��������������
This article is protected by copyright. All rights reserved.
19
The value was taken as a proxy for the difference in partners’ institutional logics. The variable
can take values from 0 to 1, where 0 represents highly compatible institutional logics and 1
represents highly incompatible logics.
To ensure the validity of the proxy for the incompatibility of institutional logics, we
computed two additional variables – CER revenues and sustainable development – and examined
the association with our proxy. As noted, we argue that CDM projects using non-renewable
technologies are more closely aligned with the market logic, because organizations engaged in
these projects seek to maximize their private benefits from obtaining CERs (Kolk and Mulder,
2011; Ashraf, Meschi and Spencer, 2014). Conversely, organizations engaged in renewable
energy projects seem willing to forego some of these private benefits. If the proxy is valid,
therefore, the revenues organizations generate from obtaining CERs should be negatively
associated with the number of renewable energy projects. We computed CER revenues by
multiplying CERs of each organization with the carbon price. Given that CER prices were not
available until 2008 when its trading started via energy exchanges, and to have carbon prices
from a single source (Bluenext – European Emissions Exchange) for the observation years 2005
to 2010, we used spot European Union Allowance (EUA) prices as a proxy for CER prices as the
former was mostly used as a base price for the latter. We found that organizations’ renewable
energy projects have a low and negative correlation with CER revenues (rho= - 0.22, p<0.01).
In addition, we argue that renewable energy projects are more closely aligned with a
public good logic, because they aim to contribute to the broader sustainable development agenda
(Geels, 2010). Accordingly, efforts to address sustainable development in CDM projects should
be positively associated with the number of renewable energy projects. To measure sustainable
development, we computed a sustainable development index by content-analyzing the collected
This article is protected by copyright. All rights reserved.
20
1353 CDM project documents to identify the meanings that actors assign to particular practices
(Reay and Jones, 2015). Using NVivo, we classified stated CDM projects’ benefits into twelve
categories: employment creation; welfare or improvement of local living conditions; learning or
facilitation of education; reduction of health risks; growth and economic development; improved
access to energy; reduction in the use of balance of payment; technology transfer; air quality
improvement; water quality improvement; avoiding soil pollution; and conservation and
protection and management of resources.5 A Cohen’s Kappa of 0.82 suggests sufficient inter-
coder reliability (Neuendorf, 2002). Following Olsen and Fenhann (2008), we then dichotomized
the categories based on whether a project had the said benefit or not. To compute a project’s
sustainable development scores, we aggregated the twelve categories’ scores6 and generated the
normalized index (cf. Barnett and Salomon, 2012):
S� =∑ ∑ "�,$$
�=1&�=1
&� ,
where S� = ith organization’s sustainable development index, "$ = nth dimension, and m = total
projects of organizations i. We found that organizations’ renewable energy projects are
positively correlated with the CDM projects’ normalized index of sustainable development
benefits (rho = 0.34, p<0.01). The correlation statistics for CER revenues and sustainable
development benefits with renewable energy projects thus both validate our proxy for the
incompatibility of logics.
Resource Dependence. There is a resource dependency between CDM project partners as
they depend on each other’s complementary skills and resources. For organizations operating in
developing countries, the incentives for reducing emissions are to gain access to technology from
developed countries’ partners, legitimacy in local and international markets, and a first-mover
advantage in becoming climate-change resilient. For developed countries’ organizations, the
This article is protected by copyright. All rights reserved.
21
incentives are to earn CERs to comply with climate regulations in their home country, to obtain a
favorable position in the carbon market, and to have new avenues to diversify their investment
portfolio (Kolk and Mulder, 2011; Ashraf, et al., 2014).
We computed resource dependencies by capturing the “dependence resulting from the
magnitude of exchange [and] fraction of business done with a partner” (Gulati and Sytch, 2007:
44). We took into account that such dependencies play out at the industry level and are
contingent on the availability of alternatives (Pfeffer and Salancik, 1978; Gulati, 1995) and
operationalized inter-industry dependence as the extent to which industries exchange with one
another (Casciaro and Piskorski, 2005). Since many organizations in the CDM market were
operating as investment vehicles of large groups and information about such vehicles was not
available, we used parent organization’s industries – four digit US standard industrial
classifications (USSIC) codes – to compute inter-industry exchange in the full dataset. The main
indicator to assess whether partners have met their objectives from the exchange relationships
within the CDM market is the number of CERs earned from working together. While we did not
have information about the direction of this inter-industry flow (in terms of purchase and sales),
it makes sense to use the magnitude of earned CERs as a reflection of the extent of one
industry’s dependence on the other.
We developed two variables, one for mutual dependence and one for power imbalance
between partners (Casciaro and Piskorksi 2005, Gulati and Sytch, 2007). To operationalize these
variables, we first computed the magnitude of exchange of industry i with j (Pij), which is high to
the extent that industry i earns CERs by working with industry j. To convert this exchange
measure into the dependence of organizations in industry i on those in industry j ('→�), we
multiplied it by the Herfindahl concentration ratio in industry j (Hj), because an industry’s
This article is protected by copyright. All rights reserved.
22
concentration level reflects the number of alternative partners available. We defined the
dependence of organizations in industry i on organizations in industry j as
'→� = ()�)(*)
where )+� = ∑ ,-./012
∑ ,-3012
; Zij is the amount of CERs earned by i and j working together on q number
of projects; and Zi is the total amount of CERs of industry i earned by working on p number of
projects. The H index is measured as ∑ &�4�
�56 , where m is the market share of organization a in
its relevant industry. When Hj is low, industry j is less concentrated and will have more
alternative partners available, which reduces the dependence of organizations in industry i on
those in j. Likewise, dependence of organizations in industry j on organizations in industry i is
defined as
'�→ = ()�)(*�),
To compute the power imbalance dimension of the dependence relationship (Casciaro and
Piskorksi, 2005), which captures whether organizations in industry i depend more on those in
industry j or vice versa, we took the difference7 of '→� and '�→:
Power imbalance ='�→+ − '+→�
To compute the mutual dependence dimension of the dependence relationship, which captures
the overall magnitude of exchange between organizations in both industries, we added up '→�
and '�→:
Mutual dependence = '→�+ '�→
To explain the two resource dependence variables intuitively, suppose organizations in
industry i generate 80% of their CERs by working with organizations in industry j, while
organizations in industry j earn 30% of their CERs by working with organizations in industry i.
As noted, resource dependencies are not only based on differences in organizations’ CERs but
This article is protected by copyright. All rights reserved.
23
also on the availability of alternative partners. If there are relatively many organizations that
operate in industry i (e.g., Hi = 0.5), while industry j is more concentrated (e.g., Hj = 0.7), the
power imbalance7 between organizations in both industries is 0.41 [(0.8 x 0.7) – (0.3 x 0.5)] and
their mutual dependence is 0.71 [(0.8 x 0.7) + (0.3 x 0.5)]. In other words, the magnitude of CER
exchange leads to a higher dependence of industry i on industry j, which is aggravated by the
relatively high concentration of industry j. Conversely, industry j depends less on industry i in
terms of CER transactions, and it has more alternative trading partners due to industry i’s lower
concentration level. Hence, there is a considerable imbalance in power which can be a source of
tension and lead to a breakup. Nonetheless, the substantial mutual dependence between both
partners can be a bonding mechanism for a continued relationship.
Control variables
Social network variables. To control for alternative explanations of partnership
dissolution, we analyzed several aspects of organizations’ social networks. First, we controlled
for asymmetries in network position. Organizations in a central position have a higher status
because they have easier access to information and resources in the network (Bothner et al.,
2012; Polidoro et al., 2011; Gulati and Gargiulo, 1999). Their higher status makes these
organizations attractive to others to have them as partners. The asymmetry in network position
will give organizations in a less central position an interest to sustain the relation with more
central ones and thus increase the chance of partnership survival. To take this into account, we
computed network position based on Bonacich’s (1987) centrality measure (Gulati and Gargiulo,
1999; Polidoro et al., 2011), according to which central organizations are those who are linked to
highly connected organizations or “who are highly regarded by highly regarded others”
This article is protected by copyright. All rights reserved.
24
(Bothner, Smith and White, 2010: 950). Following Ferriani, Cattani, and Baden-Fuller (2009),
we constructed and then transformed two-mode matrices (actors-by-projects) into one-mode8
adjacency matrices (actors-by-actors) for each observation year. Using one-mode matrices and
UCINET 6 (Borgatti, Everett, and Freeman, 2002), we computed the normalized Bonacich
centrality9 measure and took the absolute difference to compute position asymmetry between two
partners.
In addition, we controlled for network constraint asymmetry between partners. An
organization’s network is constraining if “partners are connected to each other” (Greve, Baum,
Mitsuhashi, and Rowley, 2010: 312). In a constrained network partners can share tacit
knowledge, but in a network with structural holes partners can access diverse and new
information. Burt (2005) demonstrated that structural holes across constrained groups are
beneficial in terms of access to diverse information. Constraint asymmetry is likely to increase
partnership survival because the partnership is a bonding mechanism for organizations to bridge
structural holes (Burt, 2005). To control for this asymmetry’s influence, we computed Burt’s
(1992) network constraint measure for each organization in a dyad based on one-mode matrices
(actors-by-actors) using UCINET 6 (Borgatti et al., 2002), and then took the absolute difference
to compute network constraint asymmetry between two partners.
It could also be that partners’ compatibility is a result of “social boundaries around
cliques” (Zuckerman, Kim, Ukanwa, and von Rittmann, 2003: 1047) of intermediaries who may
act as conduits of institutional logics. In the CDM market, designated operating entities (DOEs)
that provide consultancy, legal and auditing services (Kolk and Mulder, 2011) could play a role
in stimulating isomorphism (DiMaggio and Powell, 1983) and positively influencing partnership
survival. To control for this, we computed clique concentration as ℎ8� = ∑93:
;
�3
<= , where p
This article is protected by copyright. All rights reserved.
25
represents a dyad, d denotes DOEs, Wpd is the number of projects audited by d on which p
collaborated, and Np is the total number of projects on which p worked (Zuckerman et al., 2003).
To account for the propensity of actors to form partnerships because of having ties in the past,
we computed relational embeddedness which indicates “the extent to which a pair of
organizations (dyad) had direct contact with each other”10
in the year previous to the observation
year (Gulati & Gargiulo, 1999: 25). Previous ties are expected to breed trust and thus enhance
partnership survival.
Other control variables. Although we took partnerships between FPOs and NPOs as
cross-sector partnerships, we also reckoned with more detailed types of partnership ties by
making a distinction between the following categories: private firms and listed state-owned firms
(PVT); non-listed government owned and controlled enterprises (SOE); governments (GOV);
and NGOs. For this purpose we constructed a categorical variable for the type of partnership ties.
This variable takes the value 1 for a PVT-NGO partnership tie; 2 for a PVT-SOE tie; 3 for a
PVT-GOV tie; 4 for a NGO-SOE tie; 5 for a NGO-GOV tie; 6 for a SOE-GOV tie; 7 for a PVT-
PVT tie; 8 for an NGO-NGO tie; 9 for an SOE-SOE tie; and 10 for a GOV-GOV tie. We also
controlled for project size asymmetry (difference in the sum of average scale of projects in each
organization’s project portfolio), project experience asymmetry (difference of total number of
projects of partners), and registration rate asymmetry (difference in the rate of success to
register CDM projects with UNFCCC, computed by dividing the number of partners’ registered
projects out of total projects their portfolio). The latter was used to compute the inverse Mill’s
ratio to control for selectivity bias. We also computed resource commitment asymmetry
(difference in partners’ ratio of projects in which investment was made to import project-specific
This article is protected by copyright. All rights reserved.
26
technologies) and used this variable to test for potential endogeneity of mutual dependence due
to a hold-up problem (Williamson, 1979).
Furthermore, to take into account that partnership dissolution might stem from
geographic differences (Reuer and Lahiri, 2014; Mindruta et al., 2016), we controlled for
country-level effects by computing a categorical variable that captures partnerships between
different countries (Polidoro et al., 2011). It takes the value 1 for partnerships between
organizations from China and any country other than India; 2 for organizations belonging to
India and any country other than China; 3 for organizations belonging to India and China; 4 for
organizations both belonging to China; 5 for organizations both belonging to India; and 6 for
organizations belonging neither to China nor India. We also took into account whether
partnerships were between developing and developed country organizations, or not. We
computed a categorical variable developing-developed which takes the value 1 if a tie exists
between developing and developed country organizations; 2 between two developing country
organizations; and 3 between two developed country organizations. We used the UNFCCC
classification of non-annex I parties as developing countries and the remainder as developed
countries. We also computed a dummy variable for the cross-border tie, which takes the value 1
if partners in a dyad belong to different countries (regardless of the UNFCCC
developed/developing countries classification), and 0 if they have the same country-of-origin.
Model estimation
We used event history analysis to estimate the effects of incompatible institutional logics
and interdependence among partners on the dissolution of partnerships because our unit of
analysis was tie-dissolution at the project level. We used this type of analysis to account for the
This article is protected by copyright. All rights reserved.
27
fact that organizations’ withdrawal from projects might lead to a dissolution of a partnership tie
for that project only, as well as repeated dissolution events between the same set of organizations
during the observation years. The method is appropriate because it takes into account censoring
and different entry timings of subjects in the study (Polidoro et al., 2011). Survival analysis
models the hazard rate of an event, defined as the conditional probability of the event occurring
at time t given no occurrence until time t-1 (Kalbfleisch and Prentice, 2011). We chose the
accelerated failure time lognormal model for our analysis, as it assumes a flexible baseline
function, varying with time. This flexibility allowed us to incorporate monotonic and non-
monotonic baseline functions, which is appropriate in our case due to the uncertainty in the CDM
market and the two trading periods of the EU emissions trading scheme within our observation
years (Betz and Sato, 2006). In accelerated failure time models, the dependent variable –
duration of partnership until dissolution – is log-linearly related to other independent variables:
ln (T)=β0 +βi Xi,k +αµ,
where β0 is the constant, βi is the effect of independent variables on the dependent variable, and
αµ is the scale parameter of distribution (Park and Russo, 1996). In these models, a positive sign
is interpreted as an increase and a negative sign as a decrease in survival time of the partnership.
An increase in survival time decreases the hazard of dissolution while a decrease in survival time
increases the hazard of dissolution. These survival times are linked with the hazard of dissolution
(Park and Russo, 1996: 884) in lognormal distribution as
h(t) = (�@�A([(���(�)AC�)];)/4F;)
(G4H)F�[6AI((JKL(�)AC�)/F,
where φ is the cumulative distribution function for log normal distribution.
Our sample was not randomly selected. To account for potential selection bias, we
controlled for organizations’ propensity to form partnerships (Polidoro et al., 2011). We
This article is protected by copyright. All rights reserved.
28
followed a Heckman two-step procedure by including second-stage variables11
in the selection
equation, using registration rate as an instrument variable. We considered registration rate as an
instrument because it affects the propensity to form CDM project partnerships, but not the
dissolution of such partnerships. The first milestone in the CDM project cycle is to register a
project with UNFCCC. Organizations with a successful track record in this regard would make
them sought-after partners. We expect that organizations’ differential success rate to register
CDM projects would affect the formation of partnerships between project participants. However,
after the formation such asymmetric competencies of partners would not affect the continuation
of the partnership for potential registration of future projects. After going through the whole
cycle of project registration, such asymmetric competency would be redundant given the explicit
nature of project registration knowledge (Kolk and Mulder, 2011). Therefore, we do not expect
registration rate to affect partnership dissolution. Furthermore, our tests show that the instrument
variable (registration rate) did not affect the survival time or hazard of partnership dissolution
(Hamilton and Nickerson, 2003).12
We computed the inverse Mill’s ratio using a Probit model (χ2 = 2100, p < 0.01), which
predicted an organization’s probability of forming a partnership with a success rate of 69.40%
(see Table II). The resulting inverse Mill’s ratios were used as one of the regressors in all the
models. We estimated the models step-wise: we estimated model 1 with control variables,
introduced incompatibility with controls in model 2, followed by power imbalance and mutual
dependence along with the interaction effects in models 3 and 4, respectively. The robust
standard errors, adjusted for clustering at the dyads of cross-sector partners were computed
(Polidoro et al., 2011), as it allows for non-independence of observations and corrects for group-
wise heteroskedasticity (Greene, 2003). To account for the fact that actors’ ties might have
This article is protected by copyright. All rights reserved.
29
already been dissolved but were acknowledged only after being communicated officially to the
UNFCCC, we followed Greve et al. (2010) by lagging all the predictors.
---------------------------------------------
INSERT TABLE II ABOUT HERE
---------------------------------------------
RESULTS
There were significant correlations between the variables (see Table III). However, the
variance inflation factors (VIF) of all the variables were found to be less than 6, suggesting that
there was no issue of multicollinearity. In the following, we explain the results of our models
(see Table IV).
---------------------------------------------------------
INSERT TABLES III AND IV ABOUT HERE
-----------------------------------------------------------
As Table IV shows, the fitness of model 2 (AIC=2484) is better than base model 1
(AIC=2500). These results support hypothesis 1 that incompatibility (β= -0.276, p < 0.01)
decreases the survival time and increases the hazard of dissolution among cross-sector partners,
ceteris paribus. In Model 3, the interaction13
of power imbalance and incompatibility (β= -0.044,
p = 0.03) is statistically significant, suggesting a moderation effect of power imbalance. To
analyze the impact of varying degrees of power imbalance with varying levels of incompatibility,
the interaction plot was drawn (see Figure 1) at one standard deviation above and below the
mean for ease of interpretation. The interaction plot suggests that, as incompatibility increases, a
higher power imbalance between cross-sector partners decreases the survival time and increases
the hazard of the partnership’s dissolution. Low power imbalance between partners increases the
survival time and decreases the hazard of dissolution, which supports hypothesis 2.
-----------------------------------------------------
INSERT FIGURE 1 AND 2 ABOUT HERE
-----------------------------------------------------
This article is protected by copyright. All rights reserved.
30
In Model 4, the interaction of mutual dependence and incompatibility (β= 0.073, p =
0.02) is statistically significant. The interaction plot (see Figure 2) shows that, as incompatibility
increases, it increases the survival time for more mutually dependent partners, while it decreases
for less mutually dependent partners. This points out that the hazard of dissolution is high for
partners with low mutual dependence in comparison to those with high mutual dependence. This
finding lends support to hypothesis 3 that with an increase in mutual dependence the impact of
incompatible logics on the hazard of dissolution will be lower.
The results (model 5) show that there is no three-way interaction effect between
incompatibility, power imbalance and mutual dependence (β= 0.004, p > 0.1). Looking at the
controls, models 2 to 5 show that network position and constraint asymmetries positively affect
partnership survival time, as expected. Furthermore, the results show that project size asymmetry
and relational embeddedness negatively affect the survival time of the partnership. The negative
effects of project size asymmetry suggest that organizations are more likely to form and sustain
partnerships with those who share similar project size portfolios. The effects of relational
embeddedness are more counterintuitive. The findings suggest that organizations prefer to seek
and maintain new relationships in the carbon market instead of relying on partners from the
past.14
Compared to the base category, i.e. partnerships between organizations from China and
other countries, the results show positive effects of country-level dummies on survival time.
Relative to intra-border ties, cross-border ties positively affect the survival time, and this relation
is statistically significant (p < 0.01) in all the models.
This article is protected by copyright. All rights reserved.
31
Robustness Checks
To check for robustness of our results and rule out the possibility of unobserved
heterogeneity, particularly considering the fact that we did not have information about the
specific contractual terms and conditions between the partners that can be a source of bias in our
estimates, we ran frailty models (see models 6 to 9) based on gamma distribution with log-
normal baseline. The estimates of the frailty component (ϴ) with lognormal baseline models are
near zeros, suggesting that unobserved heterogeneity is not a concern in our models (2 to 5) and
our results are robust to these specifications. Furthermore, considering that mutual dependence
may also be considered as endogenous due to mutual hold-up, following Tchetgen and
colleagues (2015), we estimated models that regressed mutual dependence against survival
regression variables plus inverse Mill’s ratio and used resource commitment asymmetry15
as an
instrument. The computed residuals were then used in the survival regression as one of the
predictors. The results remained consistent.
Post-Hoc Analysis
To gain a more fine-grained insight into our hypothesized relationships, we first probed
partners’ differential sensitivity to power asymmetries. We found that when NPOs depend on
FPOs, incompatibility increases the hazard of dissolution (β=-0.344, p < 0.01). Power imbalance
(β=-0.077, p < 0.1) augments this relationship, whereas mutual dependence (β=0.098, p < 0.1)
decreases the hazard of partnership dissolution. However, these effects are not statistically
significant when FPOs depend on NPOs. This makes sense intuitively as we expect NPOs to be
more heavily depending on resources from FPOs than vice versa, because FPOs tend to have a
larger resource endowment overall.
This article is protected by copyright. All rights reserved.
32
While we focused on cross-sector partnerships between FPOs and NPOs in our main
analysis, our sample also contained projects that consisted either of FPOs or NPOs only. Since
both categories are aggregates of a diverse set of organizations – i.e., FPOs consist of private
firms as well as publicly-listed state-owned firms, while NPOs are made up of non-governmental
organizations and governments – we also expected an incompatibility of logics in intra-sector
partnerships, i.e. FPO-FPO partnerships and NPO-NPO partnerships. While both non-
governmental organizations and governments are considered NPOs, for example, their respective
behaviors are not fully guided by the same institutional logics. Moreover, with the blurring of the
sectoral boundaries (Kivleniece and Quelin, 2012; Pache and Santos, 2010; Teegen et al., 2004),
the public good and market logics do not always follow clear demarcation lines between the
sectors. Within sectors, organizations will show different shades of grey with regard to their
imprinted logics. In pursuing CDM projects, private firms could follow different logics, for
example. While we argued that firms tend to approach CDM projects from a market logic, e.g. to
obtain CERs for the private benefit of an additional revenue stream, there might be firms that
approach them from a public good logic instead when they consider their engagement in the
carbon market as an expression of their social responsibility.
Hence, we also probed whether incompatibility between intra-sector partners affects the
survival of partnership. The results show that incompatible logics indeed affect partnership
dissolution in intra-sector partnerships.16
However, we found weak support that incompatibility
increases the hazard of dissolution of intra-sector partnerships in general (β=-0.157, p < 0.1); this
effect is only highly significant in NPO-NPO partnerships (β= -0.570, p < 0.01). The moderation
effects of power imbalance and mutual dependence are not found to be statistically significant
either in NPO-NPO or FPO-FPO partnerships. This post-hoc analysis thus suggests that the NPO
This article is protected by copyright. All rights reserved.
33
category seems more heterogeneous than the FPO category in terms of institutional logics.
Besides, it is the incompatibility of logics that is creating the challenge between partners rather
than the sectoral background per se. Nonetheless, in view of the more generally significant
results of incompatibility in cross-sector partnerships compared to intra-sector partnerships, the
clash between the market and public good logics seems more evident, the more disparate the
sectoral background of the partnering organizations.
DISCUSSION AND CONCLUSION
Discussion of findings
In this study, we investigate factors that determine whether cross-sector partnerships can be
sustained for a considerable period of time or dissolve after a brief existence. In cross-sector
partnerships, organizations face the challenge of dealing with different institutional logics of
their partners, while relying on each other’s resources. On the one hand, extant research tends to
focus on organizations’ resource profiles as a basis for a partnership’s success (Gulati and Sytch,
2007); yet, an analysis of the impact of different institutional logics of partners is absent. On the
other hand, previous studies have discussed individual and organizational level mechanisms that
partners use to mutually adjust their organizational routines and expectations (Wang et al., 2008;
Katila et al., 2008; Seitanidi and Crane, 2009). However, the role of resource dependencies as a
way to resolve conflicts of competing institutional logics has not received much attention (Di
Domenico et al., 2009; Le Ber and Branzei, 2010; Wry et al., 2013). To address this gap, we
integrate elements from resource dependence theory and the institutional logics perspective to
gain a better understanding of factors that might undermine the survival of cross-sector
partnerships.
This article is protected by copyright. All rights reserved.
34
In line with previous findings, we argue that an understanding of the way in which
incompatible institutional logics manifest themselves in cross-sector partnerships is critical and
has significant implications for predicting the survival of a partnership. We specifically focus on
one of the most pertinent sources of institutional complexity in cross-sector partnerships: the
incompatibility between the public good logic and the market logic (Mars and Lounsbury, 2009).
This study’s findings support our first hypothesis indicating that an incompatibility between
institutional logics negatively affects the survival of a cross-sector partnership. Another principal
finding is that, although incompatible logics pose challenges for cross-sector partnerships, the
negative effect of incompatible logics is mitigated by the degree of resource interdependence
between organizations. While mutual dependence counteracted the impact of incompatible
logics, power imbalance further increased the adverse impact of an incompatibility. Given the
disparate moderation effects of both components of resource dependence, our findings thus lend
further support to the need to distinguish between power imbalance and mutual dependence
(Casciaro and Piskorski, 2005; Gulati and Sytch, 2007).
Overall, a cross-sector partnership’s highest probability of survival occurs when there is a
low incompatibility between partners’ institutional logics and high mutual dependence. In such
cases, partners rely on each other’s resources without being able to impose exchange conditions.
Since the tension between the institutional logics is low, partners are more open to adapt to
alternative institutional logics. However, even if logics are highly incompatible, so long as
mutual dependence is high, partners have the interest to sustain the relation because they
consider it as strategically important. High mutual dependence compels cross-sector partners to
opt for compromise to find a way to deal with competing logics. Partners will feel the need to
look for pragmatic solutions beyond any widespread belief system. In contrast, the results show
This article is protected by copyright. All rights reserved.
35
that power imbalance intensifies the adverse impact of incompatible logics on partnership
survival. This moderating effect suggests that the less dependent organization exercises its power
by attempting to impose its institutional logics on others in the partnership. Under this scenario,
the tension in the partnership is high, because each partner holds on to its norms, values, and
practices. Since the tension could become insurmountable, it will lead to a breakup of the
partnership, instead of an attempt to accommodate the institutional logics of the partners.
Theoretical implications
Theoretically, this paper contributes to an understanding of the failure of cross-sector
partnerships by integrating elements from institutional logics and resource dependence theory.
While a review of the alliance literature reveals an extensive list of reasons for the failure of
partnerships between firms (Park and Ungson, 2001), little is known about the specific reasons
why cross-sector partnerships fail. What makes cross-sector partnerships distinctive is the fact
that practices of for-profit and non-profit organizations bear the imprint of different institutional
logics (Vurro et al., 2010). The findings suggest that the incompatibility between the market
logic and public good logic of cross-sector partners negatively affects the survival of
partnerships. However, despite the detrimental impact of rivalry between institutional logics on
partnership success, mutual dependence between partners can stabilize a partnership, so long as a
balance of power is achieved between organizations. The study thus provides insight into how
resource dependencies and institutional logics interact in the context of cross-sector partnerships.
Our study also lends support to Binder (2007) who argued that organizations approach the
institutional complexity neither in a purely rational way nor by blindly following
This article is protected by copyright. All rights reserved.
36
institutionalized scripts. Rather, organizations choose a middle ground, contingent on the level of
dependency and incompatibility between institutional logics.
By bringing in insights from resource dependence theory, this study contributes to the
institutional logics literature and adds a new dimension to our understanding of the way
organizations deal with institutional complexity (Wry et al., 2013). This study highlights the
important role of resource dependence in resolving institutional complexity, accounting for the
dynamics of the material elements of institutional logics. Friedland and Alford (1991) argued
that society and social relations include both material and symbolic elements. Material aspects of
institutions refer to structures and practices; while symbolic aspects refer to ideation and
meaning (Thornton et al., 2012). The institutional logics perspective emphasizes the
interpenetration of the symbolic and material aspects of institutions. However, previous research
has tended to focus on symbolic and normative aspects of logics (Le Ber and Branzei, 2010; Di
Domenico et al., 2009). This study’s findings show that the clash between organizational values
and culture (symbolic aspect) could be better understood by considering the mutual dependency
of partners at a practice level (material aspect).
Managerial implications
The results of this study also have managerial implications. Managers who seek to engage
in cross-sector partnerships for instrumental reasons must look for partners who share their
values, culture, and logics. Partners with incongruent logics face many problems. For instance,
FPOs and NPOs tend to use different forms of control systems and performance evaluations. The
control system in NPOs may be informal with features such as tolerance for mistakes, ease of
access of employees to managers, open channels of communication and informal meetings,
This article is protected by copyright. All rights reserved.
37
consensus and participative decision-making and sharing of information between employees.
What is most important to them is to ensure that the public objective of the partnership is
achieved. FPOs are more likely to safeguard their private objectives and utilize formal control
processes for this purpose (Chenhall et al., 2010). Directors in NPOs are reluctant to implement
such formal control systems, since they find them irrelevant and impractical to accomplish their
public objective. Furthermore, exploiting such systems requires training of current members as
well as recruiting ones to ensure that the new formal controls are performed effectively in NPOs.
NPO directors resent performing such changes in the structure of the workforce (Ebrahim, 2005).
Resource dependency adds more pressure on managers either to adapt and fall in line, or to quit
the partnership. We caution that risks associated with seeking complementary resources through
cross-sector partnerships might impede long-term collaboration at the cost of reputation and
legitimacy.
Limitations and future research
Despite using an original and extensive dataset to test our hypotheses, the particular
empirical context of this paper (i.e. the carbon-offset market), a large number of Chinese
organizations,17
and the lack of contract-level information may limit the generalizability of our
findings. Furthermore, compared to other cross-sector partnerships, the rules surrounding CDM
projects are more clearly defined, as it is a legal mechanism governed by the United Nations.
Still, we expect that much of the cross-sector dynamics will be similar in partnerships that do not
operate in such a highly institutionalized environment. Nonetheless, as an avenue of further
research, it would be interesting to see if the results hold for cross-sector partnerships in other
contexts. For partnerships where organizations collaborate on a fully voluntary basis one would
This article is protected by copyright. All rights reserved.
38
expect that the lack of clearly defined rules would limit the partnership survival time. Then
again, the voluntary nature might also mean that disparate logics would form less of a problem,
as all partners would make the deliberate choice to work with organizations from different
backgrounds. Furthermore, it would be interesting to explore the effect of institutional logics on
the matching and selection of partners (Mindruta et al., 2016). While our study focused on the
dissolution of partnerships, more research is needed to understand how organizations make a
choice to partner with certain organizations and not with others.
Existing research has identified different scenarios for how competition between
incompatible logics plays out (Kraatz and Block 2008; Pache and Santos, 2010; Battilana and
Dorado, 2010; Lounsbury, 2007; Greenwood et al., 2011). As Meyer and Hollerer (2010: 1251)
argue: “There are multiple ways in which multiple logics can interact: logics may peacefully co-
exist, compete with one another, supersede each other, provide an opportunity for blending or
hybridization, or result in a compromise or temporary co-existence”. While our study provides
empirical support for a co-existence scenario, i.e. if there is a low incompatibility and high
mutual dependence; it would be interesting to examine to what extent the interaction between
resource dependence might lead to hybridization configurations (Battilana and Dorado, 2010).
We expect that in cross-sector partnerships, the higher the mutual dependence, the more likely
partners will try to find a compromise. When power imbalance and incompatibility between
logics are high, our findings suggest that the logics of the more powerful partner will supersede
those of the less powerful ones. Future research can explore the specific organizational
mechanisms that allow for hybridization of institutional logics. Finally, as our post-hoc analysis
shows, an incompatibility of logics will not only play out between FPOs and NPOs, but also
between organizations that operate within these broad categories. Future research could thus
This article is protected by copyright. All rights reserved.
39
develop more fine-grained analyses that focus on the impact of differences in logics between a
more diverse set of actors.
NOTES
1 http://pub.iges.or.jp/index.html
2 https://cdm.unfccc.int/
3 https://mintglobal.bvdinfo.com
4 http://www.stats.gov.cn/english/
5 See Olsen and Fenhann (2008) for details about the definitions of these categories.
6 Barnett and Salomon (2012) sum the KLD scores along thirteen performance criteria.
7 The magnitude of power imbalance between partners remains the same. However, its direction for organizations in
i and j will be different; both specifications of power imbalance i.e. '�→+ − '+→� and'+→� −'�→+ yield similar
results (available with authors). Considering the cross-sector and cross-border partnerships in the CDM market, the
exchange and power relations between industries can be expected to be asymmetric given their differential needs
(Kolk and Mulder, 2011; Ashraf et al., 2014). This computation of power imbalance is also in line with studies that
measure power asymmetry’s direction as well as magnitude (Gulati and Sytch, 2007).
8 Actors who were working on projects (2-model network) are assumed to have partnerships with each other (1-
mode network). The transformation into one-mode matrices was motivated by theoretical reasons to learn about
actor-level dynamics (Ferriani et al., 2009).
9 We selected the highest positive beta (default in UNICNET 6) to compute the Bonacich centrality measure. In
doing so, we followed existing approaches in the literature to capture asymmetries due to actors’ position, status and
access to information (Gulati and Gargiulo, 1999). Negative beta values capture power dependencies (Rodan, 2011),
which are already captured through our dependence measures.
10 Network constraint and relational embeddedness are related measures, but capture different aspects of a social
network. While the former captures the triadic structure i.e. whether organizations’ partners have relationships with
each other, the latter captures the dyadic structure of organizations’ direct ties with their partners.
11 Interaction effects were excluded due to collinearity. Polidoro et al. (2011) also did not include interaction effects.
This article is protected by copyright. All rights reserved.
40
12
The results of this analysis are available with the authors.
13 Results of insignificant direct effects of power imbalance and mutual dependence on the dissolution of
partnerships are omitted to conserve space.
14 Further analysis shows that relational embeddedness positively and significantly affects partnership survival in the
case of FPO-FPO partnerships only.
15 It is a weak instrument, as it may affect dissolution. However, we could not find such an effect (results available
with the author). Our results remained consistent if resource commitment is used as control variable.
16 Further analysis (using the SUEST command in STATA 14) shows that coefficients of incompatibility in both
cross-sector and intra-sector models are not statistically different (χ2= 1.44, p>0.1).
17 In our data, more ties exist between organizations that belong to countries other than China or India. At the tie
level – our unit of analysis – the sample thus captures the diversity, despite the dominance of Chinese organizations
in the carbon market. Nevertheless, we acknowledge this as one of the limitations of our study.
This article is protected by copyright. All rights reserved.
41
REFERENCES
Ashraf, N., Meschi, P. X. and Spencer, R. (2014). ‘Alliance network position, embeddedness and
effects on the carbon performance of firms in emerging economies’. Organization &
Environment, 27, 65–84.
Austin, J. E. (2000). ‘Strategic collaboration between nonprofits and business’. Nonprofit and
Voluntary Sector Quarterly, 29 (suppl 1), 69–97.
Austin, J. E. and Seitanidi, M. M. (2012). ‘Collaborative value creation: A review of partnering
between nonprofits and businesses: Part I. Value creation spectrum and collaboration
stages’. Nonprofit and Voluntary Sector Quarterly, 41, 726-58.
Barnett, M. L. and Salomon, R. M. (2012). ‘Does it pay to be really good? Addressing the shape
of the relationship between social and financial performance’. Strategic Management
Journal, 33, 1304–20.
Battilana, J. and Dorado, S. (2010). ‘Building sustainable hybrid organizations: The case of
commercial microfinance organizations’. Academy of Management Journal, 53, 1419–40.
Beckman, C. M., Haunschild, P. R. and Phillips, D. J. (2004). ‘Friends or strangers? Firm-
specific uncertainty, market uncertainty, and network partner selection’. Organization
Science, 15, 259–75.
Besharov, M. and Smith, W. (2014). ‘Multiple logics in organizations: Explaining their varied
nature and implications’. Academy of Management Review, 39, 364-81.
Betz, R. and Sato, M. (2006). ‘Emissions trading: lessons learnt from the 1st phase of the EU
ETS and prospects for the 2nd phase’. Climate Policy, 6, 351–59.
This article is protected by copyright. All rights reserved.
42
Binder, A. (2007). ‘For love and money: Organizations’ creative responses to multiple
environmental logics’. Theory and Society, 36, 547–71.
Bonacich, P. (1987). ‘Power and centrality: A family of measures’. American Journal of
Sociology, 92, 1170–82.
Borgatti, S. P., Everett, M. G. and Freeman, L. C. (2002). UCINET 6 for Windows: Software for
social network analysis (Version 6.102). Harvard, MA: Analytic Technologies.
Bothner, M. S., Smith, E. B. and White, H. C. (2010). ‘A Model of Robust Positions in Social
Networks1’. American Journal of Sociology, 116, 943–92.
Brown, L. D. and Kalegaonkar, A. (2002). ‘Support organizations and the evolution of the NGO
sector’. Nonprofit and Voluntary Sector Quarterly, 31, 231–58.
Brown, L. D. and Moore, M. H. (2001). ‘Accountability, strategy, and international
nongovernmental organizations’. Nonprofit and Voluntary Sector Quarterly, 30, 569–87.
Burt, R. S. (2005). Brokerage and closure: An introduction to social capital. Oxford: Oxford
University Press.
Candler, G. and Dumont, G. (2010). ‘A non-profit accountability framework’. Canadian Public
Administration, 53, 259–79.
Casciaro, T. and Piskorski, M. J. (2005) ‘Power imbalance, mutual dependence, and constraint
absorption: A closer look at resource dependence theory’. Administrative Science
Quarterly, 50, 167–99.
This article is protected by copyright. All rights reserved.
43
Chenhall, R. H., Hall, M. and Smith, D. (2010). ‘Social capital and management control systems:
A study of a non-government organization’. Accounting, Organizations and Society, 35,
737–56.
Croteau, D. and Hicks, L. (2003). ‘Coalition framing and the challenge of a consonant frame
pyramid: The case of a collaborative response to homelessness’. Social Problems, 50, 251–
72.
Dacin, M. T., Oliver, C. and Roy, J.-P. (2007). ‘The legitimacy of strategic alliances: An
institutional perspective’. Strategic Management Journal, 28, 169–87.
Di Domenico, M., Tracey, P. and Haugh, H. (2009). ‘The dialectic of social exchange:
Theorizing corporate—social enterprise collaboration’. Organization Studies, 30, 887–907.
DiMaggio, P. J. and Powell, W. W. (1983). ‘The iron cage revisited: Institutional isomorphism
and collective rationality in organizational fields’. American Sociological Review, 48, 147–
60.
Drumwright, M. E., Cunningham, P. H. and Berger, I. E. (2004). ‘Social alliances:
Company/nonprofit collaboration’. California Management Review, 47, 58–90.
Dunn, M. B. and Jones, C. (2010). ‘Institutional logics and institutional pluralism: The
contestation of care and science logics in medical education, 1967–2005’. Administrative
Science Quarterly, 55, 114–49.
Ebrahim, A. (2005). NGOs and organizational change: Discourse, reporting, and learning.
Cambridge: Cambridge University Press.
This article is protected by copyright. All rights reserved.
44
Ferriani, S., Cattani, G. and Baden-Fuller, C. (2009). ‘The relational antecedents of project-
entrepreneurship: Network centrality, team composition and project performance’.
Research Policy, 38, 1545–58.
Friedland, R. and Alford, R. R. (1991). ‘Bringing society back in: Symbols, practices and
institutional contradictions’. In Powell, W. W. and DiMaggio, P. J. (eds), The New
Institutionalism in Organizational Analysis. Chicago, IL: University of Chicago Press,
232-63.
Friedman, M. (1970). ‘The social responsibility of business is to increase its profits’. New York
Times Magazine, September 13, 122–24.
Geels, F. W. (2010). Ontologies, socio-technical transitions (to sustainability), and the multi-
level perspective. Research Policy, 39, 495-510.
Greene, W. H. (2003). Econometric analysis, 5th
ed. Upper Saddle River, NJ: Prentice-Hall.
Greenwood, R., Raynard, M., Kodeih, F., Micelotta, E. R. and Lounsbury, M. (2011).
‘Institutional complexity and organizational responses’. Academy of Management Annals,
5, 317–71.
Greve, H. R., Baum, J. A., Mitsuhashi, H. and Rowley, T. J. (2010). ‘Built to last but falling
apart: Cohesion, friction, and withdrawal from interfirm alliances’. Academy of
Management Journal, 53, 302–22.
Gulati, R. (1995). ‘Social structure and alliance formation patterns: A longitudinal analysis’.
Administrative Science Quarterly, 40, 619–52.
This article is protected by copyright. All rights reserved.
45
Gulati, R. and Gargiulo, M. (1999). ‘Where do interorganizational networks come from?’.
American Journal of Sociology, 104, 1439–93.
Gulati, R. and Sytch, M. (2007). ‘Dependence asymmetry and joint dependence in
interorganizational relationships: Effects of embeddedness on a manufacturer's
performance in procurement relationships’. Administrative Science Quarterly, 52, 32-69.
Hahn, T. and Pinkse, J. (2014). ‘Private environmental governance through cross-sector
partnerships: Tensions between competition and effectiveness’. Organization &
Environment, 27, 140-60.
Hamilton, B. H. and Nickerson, J. A. (2003). ‘Correcting for endogeneity in strategic
management research’. Strategic Organization, 1, 51-78.
Jensen, M. C. (2001). ‘Value maximization, stakeholder theory, and the corporate objective
function’. Journal of Applied Corporate Finance, 14, 8–21.
Kalbfleisch, J. D. and Prentice, R. L. (2011). The statistical analysis of failure time data (Vol.
360). New York: John Wiley & Sons.
Kaplan, S. (2008). ‘Framing contests: Strategy making under uncertainty’. Organization Science,
19, 729–52.
Katila, R., Rosenberger, J. D. and Eisenhardt, K. M. (2008). ‘Swimming with sharks:
Technology ventures, defense mechanisms and corporate relationships’. Administrative
Science Quarterly, 53, 295–332.
Kivleniece, I. and Quelin, B. V. (2012). ‘Creating and capturing value in public-private ties: A
private actor’s perspective’. Academy of Management Review, 37, 272–99.
This article is protected by copyright. All rights reserved.
46
Kolk, A. and Mulder, G. (2011). ‘Regulatory uncertainty and opportunity seeking: The climate
change clean development case’. California Management Review, 54, 88-106.
Kraatz, M. S. and Block, E. S. (2008). ‘Organizational implications of institutional pluralism’. In
Greenwood, R., Oliver, C. and Suddaby, R. (eds). The Sage Handbook of Organizational
Institutionalism, London: SAGE Publications Ltd, 243–75.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A. and Vishny, R. (1999). ‘The quality of
government’. Journal of Law, Economics, and Organization, 15, 222–79.
Le Ber, M. J. and Branzei, O. (2010). ‘Value frame fusion in cross sector interactions’. Journal
of Business Ethics, 94, 163–95.
Margolis, J. D. and Walsh, J. P. (2003). ‘Misery loves companies: Rethinking social initiatives
by business’. Administrative Science Quarterly, 48, 268–305.
Marquis, C., Glynn, M. A. and Davis, G. F. (2007). ‘Community isomorphism and corporate
social action’. Academy of Management Review, 32, 925–45.
Mars, M. M. and Lounsbury, M. (2009). ‘Raging against or with the private marketplace? Logic
hybridity and eco-entrepreneurship’. Journal of Management Inquiry, 18, 4–13.
Meyer, R. E. and Höllerer, M. A. (2010). ‘Meaning structures in a contested issue field: A
topographic map of shareholder value in Austria’. Academy of Management Journal, 53,
1241–62.
Mindruta, D., Moeen, M. and Agarwal, R. (2016). ‘A two-sided matching approach for partner
selection and assessing complementarities in partners' attributes in inter-firm
alliances’. Strategic Management Journal, 37, 206-31.
This article is protected by copyright. All rights reserved.
47
Moore, M. H. (2000). ‘Managing for value: Organizational strategy in for-profit, nonprofit, and
governmental organizations’. Nonprofit and Voluntary Sector Quarterly, 29 (suppl 1),
183–208.
Neuendorf, K. A. (2002). The Content Analysis Guidebook. Los Angeles, CA: Sage.
Oliver, C. (1991). ‘Strategic responses to institutional processes’. Academy of Management
Review, 16, 145–79.
Olsen, K. H. and Fenhann, J. (2008). ‘Sustainable development benefits of Clean Development
Mechanism projects: A new methodology for sustainability assessment based on text
analysis of the project design documents submitted for validation’. Energy Policy, 36,
2819–30.
Pache, A.-C. and Santos, F. (2010). ‘When worlds collide: The internal dynamics of
organizational responses to conflicting institutional demands’. Academy of Management
Review, 35, 455–76.
Park, S. H. and Russo, M. V. (1996). ‘When competition eclipses cooperation: An event history
analysis of joint venture failure’. Management science, 42, 875-90.
Park, S. H. and Ungson, G. R. (2001). ‘Interfirm rivalry and managerial complexity: A
conceptual framework of alliance failure’. Organization Science, 12, 37–53.
Pattison, P. E., Robins, G. L., Snijders, T. A. and Wang, P. (2013). ‘Conditional estimation of
exponential random graph models from snowball sampling designs’. Journal of
Mathematical Psychology, 57, 284-96.
Pfeffer, J. and Salancik, G. R. (1978). The External Control of Organizations. A Resource
This article is protected by copyright. All rights reserved.
48
Dependence Perspective. New York: Harper and Row.
Pinkse, J. and Kolk, A. (2012). ‘Addressing the climate change - sustainable development nexus:
The role of multistakeholder partnerships’. Business & Society, 51, 176-210.
Polidoro, F., Ahuja, G. and Mitchell, W. (2011). ‘When the social structure overshadows
competitive incentives: The effects of network embeddedness on joint venture
dissolution’. Academy of Management Journal, 54, 203-23.
Prashant, K. and Harbir, S. (2009). ‘Managing strategic alliances: what do we know now, and
where do we go from here?’. Academy of Management Perspectives, 23, 45–62.
Rangan, S., Samii, R. and Van Wassenhove, L. N. (2006). ‘Constructive partnerships: When
alliances between private firms and public actors can enable creative strategies’. Academy
of Management Review, 31, 738–51.
Reay, T. and Hinings, C. (2005). ‘The recomposition of an organizational field: Health care in
Alberta’. Organization Studies, 26, 351–84.
Reay, T. and Hinings, C. (2009). ‘Managing the rivalry of competing institutional logics’.
Organization Studies, 30, 629–52.
Reay, T. and Jones, C. (2015). ‘Qualitatively capturing institutional logics’. Strategic
Organization, 1476127015589981.
Reed, A. M. and Reed, D. (2009). ‘Partnerships for development: Four models of business
involvement’. Journal of Business Ethics, 90, 3–37.
Reinhardt, F. L., Stavins, R. N. and Vietor, R. H. (2008). ‘Corporate social responsibility through
an economic lens’. Review of Environmental Economics and Policy, 2, 219–39.
This article is protected by copyright. All rights reserved.
49
Reuer, J. J. and Lahiri, N. (2013). ‘Searching for alliance partners: Effects of geographic distance
on the formation of R&D collaborations’. Organization Science, 25, 283–98.
Rondinelli, D. A. and London, T. (2003). ‘How corporations and environmental groups
cooperate: Assessing cross-sector alliances and collaborations’. Academy of Management
Executive, 17, 61–76.
Sagawa, S. and Segal, E. (2000). Common interest, common good: Creating value through
business and social sector partnerships. Cambridge, MA: Harvard Business Press.
Seitanidi, M. M. and Crane, A. (2009). ‘Implementing CSR through partnerships: Understanding
the selection, design and institutionalisation of nonprofit-business partnerships’. Journal of
Business Ethics, 85, 413–29.
Selsky, J. W. and Parker, B. (2005). ‘Cross-sector partnerships to address social issues:
Challenges to theory and practice’. Journal of Management, 31, 849–73.
Tchetgen, E. J. T., Walter, S., Vansteelandt, S., Martinussen, T. and Glymour, M. (2015).
‘Instrumental variable estimation in a survival context’. Epidemiology, 26, 402-10.
Teegen, H., Doh, J. P. and Vachani, S. (2004). ‘The importance of nongovernmental
organizations (NGOs) in global governance and value creation: An international business
research agenda’. Journal of International Business Studies, 35, 463–83.
Thornton, P. H. and Ocasio, W. (1999). ‘Institutional logics and the historical contingency of
power in organizations: Executive succession in the higher education publishing industry,
1958-1990’. American Journal of Sociology, 105, 801–43.
This article is protected by copyright. All rights reserved.
50
Thornton, P. H., Ocasio, W. and Lounsbury, M. (2012). The institutional logics perspective: A
new approach to culture, structure, and process. Oxford: Oxford University Press.
Townley, B. (2002). ‘The role of competing rationalities in institutional change’. Academy of
Management Journal, 45, 163–79.
Vurro, C., Dacin, M. T. and Perrini, F. (2010). ‘Institutional antecedents of partnering for social
change: How institutional logics shape cross-sector social partnerships’. Journal of
Business Ethics, 94, 39–53.
Wang, H., Choi, J. and Li, J. (2008). ‘Too little or too much? Untangling the relationship
between corporate philanthropy and firm financial performance’. Organization Science,
19, 143–59.
Williamson, O. E. (1979). ‘Transaction-cost economics: The governance of contractual
relations’. Journal of Law & Economics, 22, 233-61.
Wry, T., Cobb, J. A. and Aldrich, H. E. (2013). ‘More than a metaphor: Assessing the historical
legacy of resource dependence and its contemporary promise as a theory of environmental
complexity’. Academy of Management Annals, 7, 441–88.
Yaziji, M. and Doh, J. (2009). NGOs and corporations: Conflict and collaboration. Cambridge:
Cambridge University Press.
Zuckerman, E. W., Kim, T.-Y., Ukanwa, K. and Rittmann, J. Von. (2003). ‘Robust Identities or
Nonentities? Typecasting in the Feature-Film Labor Market’. American Journal of
Sociology, 108, 1018–73.
This article is protected by copyright. All rights reserved.
51
Table I: Yearly Trend in Sample Statistics a
2005 2006 2007 2008 2009 2010
Total project level partnerships ties 220 612 3,952 5,916 9,041 7,064
Cross-sector ties 30 92 1071 1771 2899 2268
Intra-sector ties 190 520 2881 4145 6142 4796
Organizational level partnerships ties 218 472 2516 3621 4214 3281
Cross-sector organizational level partnership ties 30 84 712 1061 1174 941
Intra-sector partnership ties 188 388 1804 2560 3040 2340
Geographic distribution %
Indian organizations 21.51 12.56 5.57 5.45 6.56 10.33
Chinese organizations 6.45 10.76 24.86 30.16 20.93 18.63
% in the sample
Private for-profit firms 81.82 80.17 76.55 75.19 75.71 77.89
State owned enterprises 13.64 16.38 20.60 22.48 22.15 19.05
Government -ministries/departments 3.03 2.16 2.06 1.74 1.53 2.10
Non-governmental –organizations 1.52 1.29 0.79 0.58 0.61 0.96
For-profit organizations (%)
Private Firms 94.74 96.88 96.79 96.76 96.68 96.37
Listed State-owned firms 5.26 3.13 3.21 3.24 3.32 3.63
Not-for-profit organizations (%)
Government ministries/public sector organizations 66.67 62.50 72.22 75 71.43 69.56
Non-governmental- organizations 33.33 37.50 27.78 25 28.57 30.44
Renewable energy projects (%)
Biogas 6.67 20.87 9.16 6.49 6.07 9.23
Biomass 13.33 24.76 17.08 13.32 12.81 15.67
Wind Power 23.33 7.77 9.78 12.20 13.54 14.58
Hydro Power 13.33 11.65 31.83 36.94 36.69 26.89
Other renewable energies 0.00 1.94 0.78 0.69 0.60 0.75
Renewable projects of FPOs (% normalized by number of FPOs) 5 6 6 6 6 6
Renewable projects of NPOs (% normalized by number of NPOs) 20 16 16 14 14 14
Total CERs (in metric tons) 2292 5109 12000 8891 10500 7758
CER Prices (Euros, per metric ton) 22.52 17.33 0.66 0.02 13.02 14.4
a In 2010: total organizations = 1312; total projects =1500; total organizations in population=4019
This article is protected by copyright. All rights reserved.
52
Table II: Results of the Probit Model Predicting
Organization’s Probability of Forming the Partnershipa
VARIABLES Model 1 % changeb
Incompatibility -0.028(0.026) -0.010
Power imbalance -0.001(0.0004) -0.0002
Mutual dependence 0.002**(0.0004) 0.001**
Project size -0.094**(0.025) -0.033
Project experience -0.0004† (0.0002) -0.0001
Relational embeddedness -0.031**(0.004) -0.011**
Network position -0.002**(0.0003) -0.001**
Clique concentration 0.021*(0.008) 0.007
Registration rate -0.245**(0.028) -0.087**
India-other ties -0.646**(0.052) -0.187**
India- India ties -0.939**(0.223) -0.236**
Other-Other ties -0.312**(0.032) -0.115**
Cross-border ties 0.290**(0.024) 0.096*
Developing-developing ties -0.437**(0.078) -0.136*
Developed-developed ties 0.145**(0.028) 0.050†
NGO-SOE ties 1.328**(0.148) 0.488**
SOE-GOV ties 0.545**(0.035) 0.209**
SOE-SOE ties 0.853**(0.040) 0.329**
Constant -0.445**(0.037)
Log-likelihood -20386
χ2 2100
Percent correctly predicted 69.40 % ** p<0.01, * p<0.05, † p<0.10, two-tailed test a Number of observations= 34,165; Observation period: 2005 to 2010.
Robust standard errors are in parentheses. b Change in probability per unit change in predictors.
This article is protected by copyright. All rights reserved.
53
Table III: Descriptive statistics and correlationsa
# Variables Mean S.D. 1 2 3 4 5 6 7 8 9 10 11 12
1 Incompatibility (I) 0.211 0.321
2 Power imbalance (PI) 0.219 24.68 -0.006
3 Mutual dependence
(MD)
1.705 30.43 -0.01* 0.275*
4 Project size 0.219 0.33 0.297* 0.018* 0.086*
5 Project experience 25.27 41.51 0.423* -0.0004 0.013* 0.227*
6 Network position 14.97 32.43 0.264** -0.0003 0.003 0.224** 0.143**
7 Network constraint 0.179 0.286 0.332** -0.002 -0.006 0.367** 0.167** 0.158**
8 Clique concentration 0.729 0.967 -0.067** 0.006 0.006 -0.082** -0.121** -0.103** -0.148**
9 Relational
embeddedness
1.209 3.318 0.137** -0.006 0.01† -0.022** 0.115** 0.023** -0.02** -0.15**
10 Registration rate 0.174 0.322 0.229** -0.003 -0.011* 0.328** 0.129** 0.126** 0.401** -0.225** 0.035**
11 Resource commitment 0.171 0.237 0.236** 0.008 0.027** 0.369** 0.246** 0.153** 0.368** -0.099** 0.061** 0.629**
12 India-other ties 0.039 0.193 0.141** 0.0003 -0.0004 0.115** -0.014** 0.139** 0.213** -0.086** -0.012* 0.103** 0.147**
13 India-China ties 0.0001 0.011 -0.005 -0.0004 -0.0001 -0.007 -0.005 0.0002 0.009 -0.001 -0.001 -0.005 0.007 -0.002
14 China-china ties 0.001 0.024 -0.012** -0.0002 -0.001 -0.016** -0.012* -0.011* -0.014** -0.018** -0.006 0.02** -0.02** -0.005
15 India-India ties 0.005 0.073 -0.011* -0.0002 0.005 -0.023** -0.04** 0.006 0.01* -0.032** -0.008 -0.004 -0.025** -0.015**
16 Other-Other ties 0.821 0.383 -0.24** 0.01* 0.005 -0.253** -0.108** -0.146** -0.421** 0.224** 0.071** -0.325** -0.265** -0.43**
17 PVT-SOE ties 0.253 0.435 0.03** -0.005 -0.016** 0.047** 0.009* 0.044** 0.059** 0.021** -0.045** 0.054** 0.022** -0.022**
18 PVT-GOV ties 0.152 0.359 -0.106** 0.002 0.036** -0.088** -0.037** -0.069** -0.136** 0.039** -0.039** -0.144** -0.098** -0.076**
19 NGO-SOE ties 0.003 0.051 -0.004 -0.002 0.002 -0.003 -0.015** 0.009† 0.019** 0.025** 0.0003 0.025** 0.035** -0.01*
20 NGO-GOV ties 0.001 0.026 -0.009† 0.069** 0.068** 0.006 -0.002 -0.005 -0.014** -0.012* -0.007 0.01* -0.002 -0.005
21 SOE-GOV ties 0.041 0.199 -0.03** 0.022** 0.015** -0.022** -0.021** -0.019** -0.041** 0.019** -0.02** -0.059** -0.03** -0.042**
22 PVT-PVT ties 0.494 0.5 0.053** -0.007 -0.019** 0.012* 0.036** 0.005 0.042** -0.053** 0.084** 0.073** 0.054** 0.111**
23 NGO-NGO ties 0004 0.021 -0.002 -0.0002 -0.001 -0.014** -0.01* -0.003 0.003 -0.011* -0.006 0.003 0.002 -0.004
24 SOE-SOE ties 0.034 0.182 0.054** -0.003 -0.004 0.061** -0.008 0.052** 0.079** -0.016** -0.014* 0.046** 0.028** -0.038**
25 GOV-GOV ties 0.017 0.128 -0.034** -0.001 -0.006 -0.01* -0.006 -0.019** -0.038** 0.007 -0.017** -0.057** -0.021** -0.026**
26 Cross-border ties 0.877 0.329 0.118** 0.002 0.016** 0.144** 0.132** 0.062** 0.13** -0.067** 0.012* 0.094** 0.125** 0.075**
27 Developing-developing
ties
0.025 0.157 0.047** 0.001 -0.002 -0.007 -0.063** 0.081** 0.069** -0.083** 0.01* 0.022** -0.039** 0.273**
28 Developed-developed
ties
0.713 0.452 -0.281** 0.011* -0.003 -0.318** -0.187** -0.127** -0.456** 0.295** 0.012* -0.39** -0.338** -0.317**
**p<0.01, *p<0.05, †p<0.1
This article is protected by copyright. All rights reserved.
54
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
-0.0002
-0.001 -0.002
-0.023** -0.05** -0.158**
0.018** -0.014** -0.033** -0.124**
-0.004 -0.01* -0.031** 0.179** -0.247**
-0.001 -0.001 -0.004 -0.026** -0.03** -0.022**
-0.0003 -0.001 -0.002 0.012* -0.015** -0.011* -0.001
-0.002 -0.005 -0.015** 0.075** -0.121** -0.088** -0.011* -0.005
-0.01* 0.014** 0.066** -0.015** -0.575** -0.419** -0.051** -0.025** -0.205**
-0.0002 -0.0005 -0.002 0.01 -0.012* -0.009† -0.001 -0.001 -0.004 -0.021**
-0.002 0.021** -0.014** -0.143** -0.11** -0.08** -0.01* -0.005 -0.039** -0.186** -0.004
-0.001 -0.003 -0.01* 0.061** -0.076** -0.055** -0.007 -0.003 -0.027** -0.129** -0.003 -0.025**
0.004 -0.063** -0.197** -0.127** -0.001 0.099** 0.013* 0.01* 0.069** -0.131** -0.008† 0.049** 0.044**
0.065** 0.146** 0.457** -0.196** -0.068** -0.057** -0.008† -0.004 -0.034** 0.127** -0.003 -0.015** -0.021** -0.211**
-0.017** -0.037** -0.116** 0.736** -0.077** 0.206** -0.01* 0.016** 0.097** -0.104** -0.022** -0.099** 0.081** -0.143** -0.255**
This article is protected by copyright. All rights reserved.
55
TABLE IV: Lognormal Estimates of Influences on the Partnership Dissolution a
Lognormal Models Gamma Frailty Models
VARIABLES Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9
Incompatibility (I) -0.276** -0.278** -0.325** -0.326** -0.276** -0.278** -0.325** -0.326**
(0.105) (0.105) (0.107) (0.107) (0.105) (0.105) (0.107) (0.107)
Power imbalance (PI) 0.0003 0.001 0.0003 0.001
(0.0002) (0.001) (0.0002) (0.001)
I x PI -0.044* -0.072† -0.044* -0.072†
(0.019) (0.040) (0.019) (0.040)
Mutual dependence (MD) 0.001 0.001† 0.001 0.001†
(0.0005) (0.001) (0.0005) (0.001)
I x MD 0.073* 0.080* 0.073* 0.080*
(0.032) (0.040) (0.032) (0.040)
PI x MD 3.80e-07 3.80e-07
(6.58e-07) (6.58e-07)
I x PI x MD 0.004 0.004
(0.003) (0.003)
Project size -0.658** -0.639** -0.640** -0.659** -0.662** -0.639** -0.640** -0.659** -0.662**
(0.095) (0.097) (0.096) (0.097) (0.097) (0.097) (0.096) (0.097) (0.097)
Project experience 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
Network position 0.006** 0.007** 0.007** 0.007** 0.007** 0.007** 0.007** 0.007** 0.007**
(0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
Network constraint 0.618** 0.670** 0.669** 0.668** 0.664** 0.670** 0.669** 0.668** 0.664**
(0.130) (0.138) (0.137) (0.135) (0.135) (0.138) (0.137) (0.135) (0.134)
Relational embeddedness -0.124** -0.126** -0.126** -0.130** -0.131** -0.126** -0.126** -0.130** -0.131**
(0.040) (0.042) (0.042) (0.039) (0.038) (0.042) (0.042) (0.039) (0.038)
Clique concentration -0.0005 0.005 0.005 0.017 0.017 0.005 0.005 0.017 0.017
(0.047) (0.050) (0.049) (0.047) (0.046) (0.050) (0.049) (0.047) (0.046)
India-other ties 0.652* 0.663* 0.662* 0.597† 0.590† 0.663* 0.662* 0.597† 0.590†
(0.303) (0.307) (0.308) (0.311) (0.311) (0.307) (0.308) (0.311) (0.311)
India- India ties 3.837** 3.783** 3.632** 3.196** 3.219** 5.568** 5.365** 5.040** 4.923**
(0.396) (0.397) (0.399) (0.409) (0.408) (0.408) (0.407) (0.414) (0.413)
This article is protected by copyright. All rights reserved.
56
Other-Other ties 0.405** 0.391** 0.391** 0.356** 0.351** 0.391** 0.391** 0.356** 0.351**
(0.122) (0.124) (0.124) (0.125) (0.124) (0.124) (0.124) (0.125) (0.124)
NGO-SOE ties 0.366 0.423 0.428 0.502 0.512 0.423 0.428 0.502 0.512
(0.368) (0.368) (0.368) (0.362) (0.361) (0.368) (0.368) (0.362) (0.361)
SOE-GOV ties 0.205 0.227 0.232 0.248† 0.251† 0.227 0.232 0.248† 0.251†
(0.143) (0.146) (0.147) (0.145) (0.144) (0.146) (0.147) (0.145) (0.144)
Cross-border ties 0.550** 0.596** 0.597** 0.616** 0.617** 0.596** 0.597** 0.616** 0.617**
(0.139) (0.144) (0.144) (0.144) (0.143) (0.144) (0.144) (0.144) (0.143)
Developing-developing ties 0.298 0.263 0.260 0.210 0.204 0.263 0.260 0.210 0.204
(0.272) (0.274) (0.274) (0.274) (0.274) (0.274) (0.274) (0.274) (0.274)
Developed-developed ties 0.010 0.038 0.036 0.044 0.044 0.038 0.036 0.044 0.044
(0.099) (0.101) (0.101) (0.101) (0.100) (0.101) (0.101) (0.101) (0.100)
Inverse Mill's ratio 0.142 0.086 0.080 -0.025 -0.038 0.086 0.080 -0.025 -0.038
(0.272) (0.280) (0.281) (0.290) (0.288) (0.280) (0.281) (0.290) (0.288)
Constant 0.916** 0.845* 0.841* 0.724* 0.714* 0.845* 0.841* 0.724* 0.714*
(0.345) (0.352) (0.354) (0.361) (0.359) (0.352) (0.354) (0.361) (0.359)
Log-likelihood -1233 -1224 -1222 -1218 -1216 -1224 -1222 -1218 -1216
AIC 2500 2484 2484 2477 2480 2486 2486 2478 2482
Log-normal parameter (σ) 0.687 0.696 0.695 0.692 0.691 0.696 0.695 0.692 0.691
Overdispersion frailty parameter (ϴ) 1.82e-07 4.20e-07 1.54e-06 9.82e-07
** p<0.01, * p<0.05, † p<0.10, two-tailed test
Robust standard errors in parentheses
Observations = 5,735; Subjects=2897; Dyads=1172; Failure/ties = 630 b The coefficient are interpreted as time ratios that is eβ, if the value is greater than 1 it will increase the survival time of partnership and hence hazard of
dissolution will go down but value less than 1 implies survival time will decrease and hazard of dissolution will increase.
This article is protected by copyright. All rights reserved.
57
Figure 1: Effects of incompatibility on survival time of cross-sector partnership in the presence of
low and higha power imbalance among cross-sector partners
a High power imbalance is the power imbalance one standard deviation above the mean of power imbalance and low
power imbalance is power imbalance one standard deviation below the mean of power imbalance.
-10
12
3ln
(Su
rviv
al tim
e)
0 .2 .4 .6 .8 1Incompatibility
High Power Imbalance Low Power Imbalance
This article is protected by copyright. All rights reserved.
58
Figure 2: Effects of incompatibility on survival time of cross-sector partnership in the presence of
low and higha mutual dependence among cross-sector partners
a High mutual dependence is the mutual dependence one standard deviation above the mean of mutual dependence
and low mutual dependence is mutual dependence one standard deviation below the mean of mutual dependence.
-4-2
02
4ln
(Su
rviv
al tim
e)
0 .2 .4 .6 .8 1Incompatibility
High Mutual Dependence Low Mutual Dependence
This article is protected by copyright. All rights reserved.